[cjdavis618]

Well, I have finally gotten a break to try and install the 1100w model in my Grizzly SX3.

I have pulled all of the components out of the mill from Grizzly and I am pleased to report that all components fit in the factory locations without modification to the mill with the exception of the LCD module.

Button Panel, Motor, Keyboard fit directly in the factory locations.

The Factory pulley also fits on the motor, so you will instantly get more speed with changing factory components in addition to having a controller that can run from Mach3.

I am about done with the installation, and I will have the Mach control wires ran by the end of the evening. So far I have spent 2 hours installing this in the factory locations.

I will have cutting videos tomorrow afternoon at the latest.

[devers6]

I can't tell you how excited I am to hear this. Is this the Keling, Inc 1100W package? I am thinking about getting this for Christmas, and it is great to know that it seems to be designed as a direct retrofit for the SX3.

Can you give me a rough idea of the install time? I have been making some videos of upgrades I've made for my Grizzly SX3 for YouTube, and if I buy the package I'd really like to make a step-by-step HD installation video for it. I'd kind of like to figure out a time budget for how long the video would take - my recent experience works out to about 10 minutes of videography/photography time to each minute of final video. So a 2 hour install time would be dramatically different than a 12 hour install. Can you give me just a rough idea of how long it took you?

Looking forward to the cutting videos. The motor and electronics to me was the biggest problem with my SX3 - I converted mine to CNC with the CNCFusion Deluxe Kit before I even powered up the SX3, and not being able to control the spindle speed and on/off/direction was a big disappointment. (Another was that I didn't document the conversion on video as well - although I'm working to rectify this also). I've been following all your progress (as well as several other related threads) and it is great to hear about your success. Thank you for all your work on this - it is of great value to me, and I'm sure to hundreds of other SX3 owners out there.

[Flexo!]

Ok I've had a chance to test the motor.

Pretty bad news I'm afraid. Gandalf was right. The motor takes forever to recover once a load is applied.

Basically its incredibly easy to stall the motor if an instant heavy load is applied (the controller cuts out and beeps), but if a gradual load is applied, it recovers at about the same speed it originally accelerates (slowly in this case).

I connected the unit to a volt/watt meter and watched the wattage rise as the motor recovered with a load applied. Once it was at speed it was at about 1200watts at the meter, so within specs. Pretty fearsome at full power!

Running with a 0.83 reduction, so about 4977rpm at the spindle.

Excellent hardware - just needs to be reprogrammed to react faster. MUCH faster! Currently talking to Zhang as to what can be done. CN9 appears to connect to the controller chip, so looks like it could be re-flashed.

Uses this controller chip: dsPIC30F5015

[Journey_Man]

Excellent input Flexo! Maybe they (Zhang) will re-flash the Chip and do a re-call and send new boards to those that already have purchased them....

[mmprestine]

Most likely they just copied one of the Microchip application notes and did not really understand the principles behind the motor control and just took the Microchip software and made it work.

Here are some relevant application notes for the dspic:

AN1078

AN1106

AN1160

AN1208

Matt

[cjdavis618]

I have my 1100 watt motor installed but am fighting with some noise on my Index cable for RPM readings. I can cut now, but I don't have an accurate speed measurement. Not a show stopper but a irritation.

One note though, the SX3 BLDC system that comes from grizzly will stall as well under a heavy load and do the exact same thing. I would have to test my new motor before I can time it and compare. I have to say though that Flexo!'s comment is pretty disheartening. I wish I could have found all this out sooner before I got everyones hopes up.

I started this endeavor with the 2200w system which does respond slow compared to my 1100w motor, but both were on the bench and not in a machine. It will be some time before I can tear my lathe apart to finish that motor install of the 2200w.

However, I will compare the power of the sx3 size motor today however with video.

[plasmator]

Try some ferrite beads on that cable to reduce noise. Also double- check your ground path...

[cjdavis618]

I have the ones that came out of the mill from the factory that have been re-used.

I am going to pick up some shielded cable after work and the replace that. I can tell that I need it for not only the Controller but also my limit switches. They trigger randomly at times as well.

Not for long though.

[cjdavis618]

Edited.
(Reserving judgement till more tests are done.)

[cjdavis618]

I freed up the motor with the spindle tool release that comes with Grizzly sx3s and it seems to be free now with no issues.

Just for grins, I took a 1/2 coarse rougher and sent it through some 6061 at .100 doc, full width and really the motor handled as well as my Grizzly did with that amount of depth.
I'm going to continue testing to see what I can do with it. Stay tuned. I should have some videos soon.

[cjdavis618]

Ok, I am using this motor and able to send a 3/8" rougher through 6061 at 45ipm at .150doc at full width with no issues.


This is the factory pulleys getting the machine to 2200 rpm which is what my mill calc says should be max for that 200 sfpm. Mills a bit noisy with the old pulleys for testing, but it helps give a good indication of the speed and torque.

[nomedia="http://www.youtube.com/watch?v=VEMWgVfTHOU&feature=mfu_in_order&list=UL"]YouTube - Sangmutan 1100w mill Cutting[/nomedia]

[Flexo!]

Hey Chris, I too am coming to the conclusion that this motor combo should be sufficient for milling - sort of.

I think its salvation comes down to spindle bearing pre-load/ friction, and by the fact that when the motor is at full speed, its drawing 400-600 watts (depending on your setup) before you even take a cut - so its not going from 0 to 100% which would probably stall it!

An unloaded motor at 6000rpm draws 100watts or less.

Also the fact when you grab a collet chuck or spindle nose its a much larger diameter than a 10 or 8mm milling bit. I doubt I could stop a 10mm drill blank in a chuck.

That said, I still have concerns over its ability to run facing operations at a decent cut depth. The acceleration and recovery is still too slow. Found a Taiwaneese DC BLDC controller that you can set the accleration to 0.1 seconds! Impractical perhaps, but nice to have? Adlee Powertronic

[draughted]

this is a bit of a downer i have to say. i have been raving about this motor and controller.
i was able to get sangmutan to design a motor to my specs and i was thinking it was too good to be true. goes to show not all things are equal.

i am waiting for the package to come, and i have to say the service makes up for the doubt. sangmutan was able give me a motor that matches my taig spindle 10,000rpm 500watt and a 80 by 80mm motor frame.

i'll just hope and wait for my first impressions.

[cjdavis618]

That said, I still have concerns over its ability to run facing operations at a decent cut depth.

I have a 2.5" 4 flute facemill that i will test in just a few minutes. My calc says run it at .031 deep at 17ipm/776 rpm

That close enough?

[cjdavis618]

Well, here are the results.

Same 6061 block with the ipm and speed above. I can take .020 doc off per pass at 50% width without stalling but you can tell it strains the motor. If I go to .030 doc it stalls mid cut.

(Made mistake on speeds anf feeds above, I left HSS selected instead of carbide. Carbide speeds did much better. lol)

Here is a video of the same block I cut earlier. at .010,.020 and .030 deep at 46ipm

[nomedia="http://www.youtube.com/watch?v=gZXAkMZdT-A&feature=youtube_gdata"]YouTube - Sangmutan facemill test[/nomedia]



One side note on the set, These things have a connection to run either 110V or 220volt. I haven't tested this with 220volts of input yet and I can see where that would have a huge improvement on the intake of power and response. We are using a small jumper on this controller to bridge the capacitors to give it full speed at 110volts. It is possible that if we connect these controllers to 220v, that the caps can charge as they need to and supply the needed performance. It will take some time to get wire ran and connect, but I will try it. That jumper is pretty small to be powering another entire leg of the circuit.


Also, I have had several questions about how I calibrate the motor to the Mach3 speed input. I am loading a video to Youtube now that explains that as well.

[nomedia="http://www.youtube.com/watch?v=GFw2V2xDizM&feature=youtube_gdata"]YouTube - Spindle Motor Calibration[/nomedia]

[bjones]

How about a full width cut with the face mill, Looks like you were only cutting about 1 1/4 width on the left side where it bogged down big time.

[cjdavis618]

That is what block I had at the time. I continued facing it down until I hit solid block again. It takes nearly an hour to upload video from here so I am just going to give the results.

It will take a full width cut at the same speed but not more than .010 deep. But again, it bogs down too much to do a full width cut. It isn't a problem if you do a stepover of about 50% or so, but full width is a problem.

I am in the process of wiring 220v to that controller to see if that helps it at all.

[BobWarfield]

That's the dark side of a face mill--they're power suckers DeLuxe. I've stalled my 2HP spindle on my IH with a Glacern RF45. For that reason, I got in the habit of setting the HP limit in my G-Wizard calculator to be about half what the machine will handle.

G-Wizard will automatically scale back the feeds and speeds to accomodate the available HP. The other thing about setting a relatively low limit is it reminds me to take extra care with the workholding if I'm going to try to pump 2HP out of it in order to avoid flying debris!

But the bad news is this motor is putting in a pretty weak showing. If I punch up the cut that almost stalls: 0.030 depth, 50% width of cut, I'm showing only about 0.15 HP or 360 watts.

Seems like that darned motor should be good for more!

I'm wonder if it has something to do with the low rpms. For aluminum with carbide inserts on a facemill, you should be able to go faster. G-Wizard would have you running 2200 rpm and 60 ipm in 6061 with that facemill. I know you said you'd dialed up HSS, but I was unclear whether the 17ipm/776 rpm numbers were what you actually ran when it stalled?

If you ran faster, then you got more than 360 watts. Let us know the actual speeds and feeds so I can recompute. I have my fingers crossed to run the bigger motor on my lathe. If it doesn't work out, I'll go back to my VFD ways.

Cheers,

BW

[cjdavis618]

I was running 2200rpm at 46ipm in the videos. At depths of .010,.020 and .030. 2.5" face mill

I did run the 17ipm at 776rpm and it stalled as well doing .030 But Carbide speeds (And tool) worked much better. This is the only Face mill I have but I do know that the Grizzly motor never stalled doing this. This is for sure, not as powerful as that motor. (Which really stinks)


To add,

I use an app that came with my Alibre called Machinist Toolbox. It is telling me that to do a full width cut, at 2200rpm and 46ipm, that I need 1.6 HP to do that. So this motor is not that far off thier rating, but just not at it either.

[Flexo!]

Cheers for the videos, good to see it can manage a little facing. You can really hear the motor recovering after getting bogged down.

I'm on 240v power here in New Zealand, and it still suffers from the same symptoms. Running on 110v should just require double the amps, so should work the same regardless?

My plan at the moment (if Zhang dosent come up with a solution), is to grab the code off the chip using one of these: iCP02 USB Microchip PIC/dsPIC Programmer and posting a request (maybe even offer some $$$) in the PIC Programming section of this site - provided I cant figure it out first. I have no PIC programming experience, but I'm sure I can muddle through. Hopefully only have to modify 2-3 numbers here and there. Played around with an Arduino, so I understand the concepts.

May even be possible to re-program the masses of Sieg SX3 controllers if they use a similar controller design.

I think its unlikely that there'll be a recall, seeing they do work. We really shouldn't be doing Sangmutans work for them, though!